Study on gas migration behavior through bentonite …eaform2017.aesj.or.jp/file/PapersList/Session5/(5B-1)_T...Fig. 1 Overview of RWMC’s study on Gas migration Fig. 2 Preferential

Study on gas migration behavior through bentonite buffer material

Tomoyuki SHIMURA*, Shinichi TAKAHASI*, Masanobu NISHIMURA*,

Kazumasa KOGA**, and Hitoshi OWADA**

* Obayashi Corporation, ** Radioactive Waste Management Funding and Research Center

* 2-15-2, Konan, Minato-ku, Tokyo, 108-8502, Japan.

** 6-4, Akashicho, Chuo-ku, Tokyo, 104-0044, Japan.

Key words; Radioactive waste disposal, bentonite buffer, gas migration, performance assessment

Abstract

Gasses generated due to the corrosion of metals contained in the waste packages and engineered barrier materials

will accumulate in the Engineered Barrier System (EBS) of a radioactive waste repository, and may affect

long-term stabilities of the EBS. To evaluate gas migration behavior through the bentonite buffer materials in

conjunction with a reasonable design model is one of the challenges to be solved for the performance assessment

of a radioactive waste repository.

A series of studies have been conducted for more than 10 years under the leadership of the Radioactive Waste

Management Funding and Research Center (RWMC) in Japan to better understand gas migration behavior

through bentonite. In this report, the outline of the framework of the gas migration study by RWMC as well as

some of the results from laboratory gas injection tests are presented.

The laboratory test results focusing on the interfaces in bentonite materials (bentonite/bentonite interfaces)

showed that water flow through the interface was blocked (self-healed) due to the swelling of the bentonite. In

addition, the gas injection test after full saturation, indicated no significant differences in breakthrough pressure

compared to the intact (e.g. no interface) specimen. These findings suggest that the interface is unlikely to

become a dominant gas migration pathway in a saturated buffer.

It is expected that all findings acquired through this decade of study are summarized and integrated within this

fiscal year (2017) taking into consideration the parallel studies on gas scenario development and integration of

coupled modelling, and contribute to a reliable advanced repository design and performance assessment.

1. IntroductionThe long-term stabilities of radioactive waste

repository are generally secured by a multibarrier

system consisting of Natural Barrier and Engineered

Barrier System (EBS). A certain amount of gas will

be generated by corrosion of metals contained in the

waste packages and EBS materials. Such gases will

accumulate in the EBS of a radioactive waste

repository and may affect the long-term stability of

the EBS. To evaluate gas migration behavior through

the bentonite buffer materials, in conjunction with a

reasonable design model, is one of the challenges to

be solved for the performance assessment of a

radioactive waste repository [1].

Extensive studies on gas migration behavior have

been carried out. One of the famous international

study is the FORGE (Fate Of Repository GasEs)

project funded by EC. The following common

challenges for the understanding of gas migration

properties have been identified through the FORGE

project [2]:

- Treatment of anisotropic micro phenomena within

a macro model.

- Treatment of coupled phenomena considering both

EBS stability and nuclide migration over time.

- Up scaling from laboratory to realistic scale with

regard to phenomena understanding.

Based on the issues identified from the findings of

the in-situ large-scale Gas Migration Test (GMT,

1997~2007) at Grimsel Test Site (GTS) [3], RWMC

has conducted a series of studies from 2008 to better

understand gas migration behavior through bentonite.

In order to find solutions to the internationally

identified issues with regard to gas migration in EBS,

research and development focused on the following

three items: obtaining better date through laboratory

testing, advancement of analysis tools and scenario

development [4].

In this report, the outline of the framework of the

gas migration study by RWMC, as well as some of

the results from laboratory saturation and gas

injection tests, are presented.

6th East Asia Forum on Radwaste Management Conference November 27-29, 2017, Osaka, Japan

Fig. 1 Overview of RWMC’s study on Gas migration

Fig. 2 Preferential gas pathways due to interfaces

2. Studies on gas migration by RWMC The studies on gas issues, e.g. gas generation, gas

migration in EBS, and effect on long-time repository

stabilities, are being conducted by JAEA and NUMO

for Japanese HLW and Transuranic (TRU) waste

repositories [5], [6], [7]. Especially in Japanese TRU

waste repository, the following phenomena or critical

events have been considered in migration of gases

generated in the waste packages, that could

significantly influence the performance of the

disposal system:

- Gas transport and diffusion to the outside of the

EBS in the form of dissolved gas in groundwater.

- Gas pressure pushing contaminated water out from

the waste package and disposal cavern into the

geosphere.

- Excess gas/water pressure causing stress on the

EBS components leading to mechanical failures

and preferential pathway creation, allowing

gas/water to seep out.

These phenomena have been considered in the

groundwater advection scenario as increased nuclide

migration due to the expulsion of contaminated water

resulting from gas generation in the EBS. The

performance assessment has been conducted with the

basic expulsion model. It showed that the maximum

dose calculated was not significantly different from

the results of the reference case.

Due to the background mentioned above, the still

ongoing study on gas migration in Japanese TRU

waste repository was initiated in 2008 by RWMC

with the purpose of reducing the uncertainties in the

performance assessment of a multibarrier system.

Fig. 1 shows the interrelationship of each activity of

RWMC’s gas study project.

RWMC’s study on gas has been ongoing for over

ten years in two phases. In the first phase

(2008~2012), studies on gas migration behavior were

conducted that take into consideration the issues

identified in the TRU 2nd progress report [6],

specifically laboratory gas injection tests for

understanding gas/water permeability in bentonite

material, development of integrated modeling and

analysis method, and scenario assessment with

consideration of the transition of repository. While

substantive findings relevant to gas migration in

bentonite were acquired through the first research

phase, several issues, which arose from uncertainties

of gas behavior, remained.

In the subsequent second phase (2013~2017),

laboratory tests targeted at not only bentonite

materials but also cementitious materials have

proceeded to enhance data related to gas/water

permeability, which contribute to reduce the

uncertainties of gas migration behavior. In addition,

modeling approaches to apply the parameters

estimated from laboratory test results have also been

investigated to trace laboratory test sequence with

higher reliability.

Through a series of studies, it has been foreseen

that the interfaces between bentonite materials may

induce preferential pathways for gas generated in a

waste package (cf. Fig. 2). RWMC has therefore


Injection line (gas)Load cell

Specimen

Outflow (gas/water)

Injection line (water)

Central OutsidePartition O-ring

Porous metals

Stainless column

O-rings

Fig. 3 column test equipment

Tab. 1 Specification of specimen

Interface (w=1mm)

Bentonite specimen (Φ60mm)

Before water injection

After saturation

Fig. 4 Bentonite column with interface Fig. 5 Flow chart of test procedure

initiated a laboratory gas injection test program to

evaluate the effect of interfaces (bentonite/bentonite

interfaces), which are expected to exist in bentonite

materials, on gas migration behavior. The gas

injection tests have focused on three phenomena,

(re-)saturation, gas injection, and gas breakthrough

(sudden flow increase), in a potential gas migration

scenario.

The latest findings through the gas injection test

will be discussed in the next chapter.

3. Laboratory tests focusing on the existence of

interface in bentonite

(1) General idea and condition of the test

The laboratory equipment consisting of gas/water

injection system, gas/water outflow measurement

system, test column, and operation panel was set up

prior to initiation of the gas injection test. Two types

of Na bentonite (Kuni-gel V1) column specimens,

with and without interface were prepared respectively.

Both specimens had a diameter of 60mm and a height

of 25mm and were placed into the column of the test

equipment as shown in Fig. 3. The specimen was

compression-molded as two layers to reach the

required dry density. The specimen without interface

was named Case-1 and the one with interface was

named Case-2. With respect to the initial condition of

both specimens, the dry density of 1.36 Mg/m3 was

set as referred to in the concept of the TRU 2nd

progress report. The water content of 32.4%

(saturation ratio of ca. 90%) was set to reduce the

duration of saturation (cf. Tab. 1). The specimen was

compression-molded statically (compression speed of

1mm/min) by two layers to reach the required dry

density. As for the Case-2, a plate of 1mm thickness

was set at the center of the column container which

was removed after compression.

The (re-)saturation phase was initiated by water

injection from the bottom of the column. At the

beginning of water injection, no water pressure was

applied to prevent the appearance of flow paths at the

outer peripheral surface of the column specimen, and

saturation proceeded only by capillary pressure of the

bentonite material. After confirmation of stable water

injection, the injection pressure was increased to 0.2


MPa which was lower than the expected swelling

pressure (ca. 0.5 MPa) of the bentonite specimen.

Full saturation of the specimen was comprehensively

assessed by multiple factors such as water seepage

from the top surface, total volume of injected water,

flow balance of inflow and outflow, water

permeability, and pH. Two fully saturation specimens

were made for Case-2 so that the actual saturation

ratio could be assessed by dismantling one specimen

before proceeding to the gas injection phase. Fig. 4

shows the top view of the Case-2 specimen: before

initiating water injection (top), and after saturation

(bottom). The interface of the bentonite self-healed

due to the swelling.

After confirming saturation of the bentonite

specimen, the gas injection phase was initiated by

closing the water injection valve and opening the gas

injection valve, both set at the bottom of the column

specimen. Nitrogen gas was injected under a

backpressure of 0.1 MPa after which the gas injection

pressure was increased by stepwise pressurization of

0.1 MPa per two days until gas breakthrough (sudden

flow increase) occurred. The temperature of

laboratory was kept at 25 degree centigrade during all

test sequences.

Fig. 5 shows the flow chart of the test procedure.

(2) Test results of Case-1: without interface

As test result of the (re-)saturation phase of Case-1,

cumulative amount of water inflow, water injection

pressure, and swelling pressure during the water

injection are summarized in Fig. 6. In the first 3 days

of the water injection, the saturation proceeded only

by capillary pressure of the bentonite as there was no

water pressurization. After that water injection

commenced at 0.2 MPa injection pressure.

Water seeping from the top surface was observed

after 36 days. The cumulative amount of injected

water at that time was around 8.4mL which was

reasonable considering the volume of the water

needed for saturation of the specimen (3.49 ml) and

void volume of the porous metal set onto the top

surface of the specimen (3.86 ml). The water

permeability determined was 6.6x10-15 m/s, which is

a lower value compared to past research for the 1.36

Mg/m3 dry density of bentonite (10-11~10-13 m/s). The

pH value of the outflow was weak alkaline at around

8 which suggested that the water came from the

bentonite specimen. Based on a comprehensive

assessment of these results, the water injection /

saturation phase was finished. As Fig. 6 indicates, the

swelling pressure measured during the test was 0.47

MPa.

Following the saturation phase, the nitrogen gas

injection phase was initiated. A back pressure of 0.1

MPa was applied to the specimen in accordance with

the test procedure, then the gas injection increased

stepwise by 0.1 MPa per two days. Fig. 7 shows the

change of effective gas pressure and gas inflow

during gas injection. As result of the gas injection

increase of 0.1 MPa per two days, the gas

breakthrough with sudden flow increase occurred

after 29 days, when gas injection pressure reached 1.6

MPa. Gas injection was automatically stopped after

the gas flow rate reached 1,000 Nml/min due to the

operation of a safety valve, at which time the whole

test sequence of Case-1 was completed.

(3) Test results of Case-2: with interface

As test result of the (re-)saturation phase of Case-2,

which included an interface at the center part of the

column specimen, cumulative amount of water

inflow, water injection pressure, and swelling

pressure during the water injection are summarized in

Fig. 8, as with the Case-1 results. In Fig. 8, the

swelling pressure was calculated by subtracting water

pressure from the total pressure measured at the

bottom. After 42 days from the start of water

injection, water seeping from the top surface was

observed and it was confirmed that the total amount

of inflow and outflow were consistent. After filling

the outflow piping with water, the measured inflow

rate and outflow rate were balanced to determine the

permeability of the bentonite specimen. The water

permeability estimated was on the order of 10-13 m/s

which is consistent with past research results. Based

on a comprehensive assessment of these results, the

saturation phase was finished after 100days of water

injection at which time the observed swelling

pressure was 0.33 MPa. The swelling pressure was

within the variation range of past research [8]

although the pressure was slightly lower than for

Case-1 (no interface).

The gas injection phase was initiated after

confirming the saturation by the same procedure

(nitrogen gas injection, and pressurization of 0.1 MPa

per two days) as for Case-1. The change of effective

gas pressure, water outflow from the top surface, and

effective stress at the column bottom during gas

injection phase are shown in Fig. 9. As the result of

the stepwise increase of gas injection, effective stress

at the bottom gradually decreased accompanying the

stepwise pressurization. Gas breakthrough occurred

after 25 days when gas injection pressure reached 1.3

MPa, which is slightly lower than observed for

Case-1. The gas injection was stopped afterwards and

the test specimen was dismantled to investigate the

inside of bentonite material.

Fig. 10 shows the distribution of water content

inside the Case-2 column specimen before gas

injection (left) and after gas breakthrough (right).

Both profiles were measured from sliced fragments

by dismantling as two saturated specimens were

prepared. The water content before gas injection

decreased from the bottom toward the top of the

column. Moreover, it was confirmed that the water


Fig. 6 Saturation phase (Case-1) Fig.7 Gas injection phase (Case-1)

Fig. 8 Saturation phase (Case-2) Fig.9 Gas injection phase (Case-2)

Fig.10 Measurement of water content inside the bentonite

Case-2, after saturation (left), and after breakthrough (right))

content around the interface is slightly higher,

compared to other areas. One of the reason for the

variation of water content from the bottom to the top

is thought to be due to the time lag of swelling and

consolidation in terms of places and time with the

water injection from the bottom to top. The time lag

of swelling causes spatial differences of density and

the water content will change accordingly. Similar

phenomena are likely to occur around the gap 1mm

wide. In the profile of the after gas breakthrough

sample, a decrease of water content from the top

toward the bottom is observed, but no significant

differences are seen between areas in the vicinity of

the gap and other areas. One of the reason of this

phenomena is thought to be that the gas progressed

gradually inside of the specimen from the bottom and

pushed out the pore water at the same time.

Continuation of gas flow after gas breakthrough


could also be a factor reducing the water content.

Meanwhile, the water content at the top surface of the

specimen before gas injection and after breakthrough

show almost identical values. This could be because

the gas breakthrough occurred before the gas front

reached the top surface of the specimen, although it

must be noted that both specimens were not the same.

(4) Lessons Learned from laboratory tests

The gas breakthrough pressure obtained through

gas injection test was 1.6 MPa in Case-1 (no

interface) and 1.3 MPa in Case-2 (with interface).

Both pressures were slightly different but reasonable

well within the range of variation observed in past

studies. Therefore, it is thought that the existence of

the interface has no effect on the gas breakthrough

pressure although this is based on the limited results

with a 1mm wide interface. The profile of water

content after gas breakthrough does not show obvious

differences between interface vicinity and other areas

and this means that the interface does not function as

a preferential gas pathway. In addition, it is

recognized that the interface may affect the

distribution of water content and dry density in an

extremely small range.

As 1mm width of the interface was chosen for this

study, the effect of different widths would be an issue

to be confirmed in further studies.

4. Conclusions As the latest findings through the RWMC’s study

on gas migration behavior which proceeded over ten

years, the laboratory gas injection test to evaluate the

effect of interfaces expected in bentonite material

was discussed in this paper.

The major findings of this study are:

- The existence of interface has no significant effect

on the gas breakthrough pressure.

- The interface does not play a role of a preferential

gas pathway.

- The interface may affect the distribution of the

water content and dry density in an extremely

small range.

- The effect of the width of interfaces would be an

issue to be confirmed in future studies.

Further challenging test program focusing on the

width of interface and the interface between different

materials are proceeding under the direction of

RWMC.

It is expected that all findings acquired through this

decade of study are summarized and integrated

within this fiscal year (2017) taking into

consideration the parallel studies on gas scenario

development and integration of coupled modeling,

and contribute to the reliable advanced repository

design and performance assessment.

Acknowledgement

This research was initiated within a project to

develop Geological Disposal Technologies in Japan,

which was funded by the Ministry of Economy,

Trade and Industry (METI), Japan.

References [1] Namiki. K, Asano. H, Takahashi. S, Shimura. T,

Hirota. K (2014): Laboratory gas injection tests

of compacted bentonite buffer material for TRU

waste disposal, Geological Society London,

Special Publications, first published May 12,

2014.

[2] Shaw. RP (2014): Overview and Key

Achievements of the FORGE Project., FORGE

Report D0.07-R.

[3] Shimura. T, Vomvoris. S, Lanyon, G. W.,

Marschall. P, Ando. K, Yamamoto. S, Asano. H

(2008): Gas Migration through Bentonite/Sand:

Lessons Learned from the GMT in situ Test,

Science & Technology Series n° 334, Andra.

[4] Koga. K, Owada. H (2016): Study on coupled

gas migration behavior in TRU disposal facility

(1) -Outline of project-. Proceedings of 71st

JSCE Annual Meeting, CS13-040 (in Japanese)

[5] JNC (Japan Nuclear Cycle Development

Institute) (2000): H12: Project to Establish the

Scientific and Technical Basis for High-Level

Radioactive Waste Disposal in Japan: Second

Progres Report on Research and Development

for the Geological Disposal of High-Level

Radioactive Waste in Japan, Supporting Report

2, Repository Design and Engineering

Technologies, JNC TN1410 2000-003.

[6] JNC (Japan Nuclear Cycle Development Institute)

and FEPC (Federation of Electric Power Companies

of Japan) (2005): Second Progress Report on

Research and Development for TRU Waste

Disposal in Japan, JNC TY1400 2005-013,

FEPC TRU TR 2-2005-02.

[7] NUMO (Nuclear Waste Management Organization

of Japan) (2013): A study on improving the

safety of TRU waste disposal in Japan -Study on

disposal system design aimed at improving

disposal safety and the reliability of the

assessment of TRU waste disposal for a range of

geological conditions-, NUMO TR14-03.

[8] Fujiyama. T, Tanaka. Y, Hironaga. M, Mori. A

(2008): Literature survey on evaluation of gas

migration through bentonite and cementitious

materials., CRIEPI Research Report, N07021.


Study on gas migration behavior through bentonite …eaform2017.aesj.or.jp/file/PapersList/Session5/(5B-1)_T...Fig. 1 Overview of RWMC’s study on Gas migration Fig. 2 Preferential

Documents